Multimodal Model Training and Inference Optimization Engineer

Work on ByteDance’s Vision-Applied Research team, optimizing multimodal generative AI models for faster, more scalable training and inference. Improve large-model training pipelines, implement distributed training strategies (data/model/pipeline parallelism and communication), and benchmark/profile deep learning workloads to find bottlenecks. Apply strong Python/C++/CUDA and deep learning framework expertise (PyTorch, Megatron, DeepSpeed) to accelerate deployment of large-scale models.